REXX has been around on the iSeries since, well, before my time. But rarely have I seen it utilized. However, I've found some simple uses for it that I'd like to share.

Use REXX as "Source" for Non-Source Objects

Often, iSeries applications consist of objects that have no source to define them. Instead, they're created with CL commands. Such objects may include data areas, data queues, and the like. And now, with ILE, service programs and binding directories also fall into this category.

It would be nice to see the CL commands that created these objects easily, without having to wade through the output of DSPwhatever commands. You could put those commands in CL programs, but then you'd have additional objects. This is why REXX is ideal for storing Create (CRTxxx) commands for your non-source objects and even for running the Create commands, without the need of a command line and prompting.

When you create these "source" members, you can and probably should set the source member type to reflect the object you are creating. If you do, remember that PDM Option 16 and RSE Run Procedure won't work, but you can easily place STRREXPRC in your own PDM option or RSE User Action.

Here's an example for a data area:

Here's an example for a binding directory:

Use REXX for One-Time and Fix-It Programs

The nice thing about using REXX for one-timer solutions is that, as I said before, there is no program object. Also, you can easily mix CL and SQL in the same REXX procedure (something that cannot easily be done in CL).

Here's an example:

And here's another:

Use REXX for Oft-Used Commands, Especially if They Are Complex

I once worked with an ODBC product that involved incredibly complex OS/400 commands. The commands were ugly. Even when these commands were put into CL programs, the commands were still ugly and hard to maintain. Using REXX allowed me to code these commands in a readable, maintainable format. Also, I did not have to recompile a CL program whenever a change was necessary.

Here's an example, using that ODBC product:

What You Need to Know

You do not need to be an expert in REXX to use it for these functions. By looking at the examples above, you can see how simple it can be. You only need to know these basics:
Begin a REXX procedure with a comment. Comments in REXX are like those in CL; they're embedded between /* and */.

• There is no need to declare variables in REXX. Simply assign a value to them.
• You can enclose the text of your command or SQL statement in single or double quotes. I suggest double quotes, in case something in the text contains single quotes.
• To put the contents of a variable into the command or SQL text, close the text with the quote, put in the variable, and use the quote to open any remaining text. If you do not want a space between the text and the variable contents, abut the text and the variable. If you do want the space, put a space between the text and the variable.
• If you need to continue a command or SQL statement to the next line, type a comma at the end of the previous line. Put a space before the comma if you need a space at that point of the command or SQL statement.
• Precede an SQL statement with address execsql execsql. Be aware of your commit level. If the file is not journaled, set the SQL commit option to *NONE, or use WITH NONE on your DML statements.
• Statements in a simple REXX procedure may end with a semicolon. It is optional.

With all the other technologies available on the iSeries, REXX still has some uses. Avail yourself of REXX and make your life easier.

Doug Eckersley is the iSeries programmer with a premier homebuilder in Columbus. He has been programming on the iSeries for 10 years and has been in the business for 15. He is certified by IBM.